Instant answers & knowledge cards#

These two endpoints mirror the website’s inline answers and side panels. Because their content varies by answer type and provider, they return provider-shaped JSON rather than a fixed per-field schema, treat the structures as data and read the fields you need.

Instant answers#

GET /api/v1/instant/?q=...

Returns the instant answer for a utility query (maths, unit/base/colour conversion, weather, currency, world clock, hashes, QR, “what’s my IP”, …), or null when the query isn’t one Seurch answers inline.

curl -H "Authorization: Api-Key seurch_sk_<prefix>.<secret>" \
  "https://search.example.com/api/v1/instant/?q=100+usd+to+eur"

{
  "query": "100 usd to eur",
  "answer": {
    "type": "currency",
    "...": "..."
  }
}

When there’s no instant answer:

{ "query": "some random phrase", "answer": null }

The answer object’s shape depends on the answer type (currency, weather, calculator, unit, …). See the user-facing list for what can match, and note that triggers are multilingual.

Knowledge cards#

GET /api/v1/cards/?q=...

Returns the knowledge panel for a query, derived from the same web + Wikipedia context the website uses. Accepts lang, safe and engine.

{
  "query": "inception",
  "wikipedia": { "...": "..." },
  "thetvdb": { "...": "..." },
  "tripadvisor": null,
  "stackexchange": null,
  "map": null
}

Each card is null when it doesn’t apply to the query (only the relevant cards are populated, the same logic as the website, so a film query fills thetvdb, a place query fills tripadvisor and map, and so on). Card availability also depends on which providers the instance has configured.

The tripadvisor card carries name, geo, address, latitude, longitude, rating, rating_bubbles, num_reviews, description and url. It no longer carries photo, location_type, cuisine, subcategory, price_level or ranking: those came from TripAdvisor’s retired Content API, and its replacement only serves them to individually licensed partners.

A single cards/ request runs a web search plus a Wikipedia lookup behind the scenes (to detect the subject), so it counts as one search toward your usage.